A novel technology that enables drones to experience pain could prevent your self-driving car from causing harm to itself.

      Drones are the first innovation, followed by autonomous vehicles. A system that senses pain and detects failures before they occur has significant implications for self-driving cars.

      When you twist your ankle while running, your body sends a pain signal to your brain, prompting you to stop. This ability to feel pain serves to prevent further injury.

      Researchers at Delft University of Technology and Wageningen University have implemented this idea in drones, creating a digital version of a nervous system that can identify a malfunctioning component and emits a pain-like warning signal. This technology could also be beneficial for autonomous vehicles.

      So, how does this "pain" system function?

      The research team created early warning indicators, termed “critical slowing down” signals, which are based on a concept previously used to foresee ecosystem collapses in ecology. Their findings were published in the Proceedings of the National Academy of Sciences (via TechXplore).

      Any intricate system, whether biological or mechanical, begins to show minor changes in its sensor data prior to actual failure. This system detects such changes using only real-time data, without the need for predictive models or historical benchmarks.

      They experimented with quadrotors at the CyberZoo drone research center by systematically damaging rotor blades, reaching up to 55% tip damage. In their experiments, loss of control was observed at 15% tip damage on the front-right rotor, and the system successfully identified the growing instability.

      “Our approach is comparable to how humans experience pain,” stated lead researcher Jasper van Beers. “After an injury occurs, pain provides immediate feedback about our state and helps us determine what actions are still safe. Typically, machines lack this kind of self-awareness.”

      How could this benefit your vehicle?

      This same principle applies to autonomous cars and advanced driver-assistance systems, particularly those used commercially as robotaxis.

      A self-driving vehicle facing a failing sensor, a malfunctioning actuator, or challenging road conditions that push it to its handling limits encounters similar issues. It cannot sense a warning before losing control.

      Since the system relies solely on real-time data, it does not need any modifications or new hardware; it uses the existing setup. The researchers specifically mention applying this technology to self-driving vehicles, which sounds particularly intriguing.

      Over the past five years, Shikhar has continually simplified developments in consumer technology.

      A counterfeit Maccy clipboard app is circulating, which is aimed at stealing passwords.

      PamStealer malware is masquerading as Maccy to target Mac users.

      A fraudulent version of Maccy, a well-known clipboard manager for macOS, is being used to deliver a recently identified strain of Mac malware called PamStealer. Researchers at Jamf report that this malware pretends to be the genuine open-source app but is actually designed to steal sensitive data and capture users’ login credentials. PamStealer is delivered as a disk image containing an AppleScript file that mimics Maccy. When a user opens this file, macOS runs it in Script Editor, where the displayed instructions prompt them to press Command-R. This may appear unusual to someone expecting a standard application installation process, but this action executes hidden malware code and initiates the attack.

      Claude Fable 5 is discontinuing subscriptions, though it may not be permanent.

      High demand is leading Claude Fable 5 to step back from subscriptions for the time being.

      Anthropic’s most advanced Claude model available to the public is set to discontinue standard subscription access after July 7, but the company is reassuring users that this change may not be lasting. Fable 5 has recently returned to Claude after facing scrutiny from the U.S. government. Anthropic stated that it would be included in Pro, Max, Team, and select Enterprise plans for up to 50% of weekly usage limits until July 7. Following that date, the model will shift to usage-credit billing, which means users will incur charges for access outside their regular plan limits.

      Another study highlights serious security flaws in AI browsers, bursting the hype bubble.

      Four popular AI browsers are vulnerable to exploitation, allowing attackers to steal data from other open tabs.

      AI browsers are being marketed as a breakthrough technology. They can summarize web pages, help book travel, and even make purchases on your behalf. However, a recent study from the University of Washington found that four out of the seven most popular AI browsers present security risks significant enough to enable malicious sites to capture data from other active tabs. The more advanced the browser functions, the greater the associated risk.